Joint inversion of Rayleigh and Love wave dispersion curves for improving the accuracy of near-surface S-wave velocities

Abstract

High-frequency surface-wave techniques are extensively employed for the determination of S-wave velocities of near-surface materials. The absence of the measured phase velocity dispersion curves at specific frequency band, however, is common in the real world, which may cause inaccurate S-wave velocity by inversion. Additionally, surface-wave phase velocities are relatively insensitive to S-wave velocities under the abnormal high or low velocity layer, so it is difficult for surface-wave techniques to obtain S-wave velocities in irregular models which contain an (or several) abnormal layer(s). Based on significantly different sensitivity frequency bands of Rayleigh and Love wave phase velocities, a joint inversion method of Rayleigh and Love wave dispersion curves was proposed by the conjugate gradient algorithm. We used synthetic models to test the effectiveness of the developed method. Numerical modeling results revealed that: (1) S-wave velocities acquired by joint inversion of Rayleigh and Love wave phase velocities were more accurate than that by individual inversion in the case of lacking surface-wave dispersion data; (2) joint inversion efficiently improved the accuracy of S-wave velocities of irregular models. A real-world example verified the validity in application of the proposed approach.